The present invention is related to a data backup and recovery system of a computer system having a computer and a storage.
Along with the increase in data capacity of storages in a computer systems of a business organization or the like, not only the data used for normal operations but also the backup data of the data used for normal operations have increased. Backup data is acquired routinely at certain intervals (for example, once a day), and it is common to acquire multiple generations of backup data corresponding to the data for normal operation. If a failure occurs during normal operation, the data at a certain point in time is recovered using the backup data. However, even if multiple generations of backup data are acquired, the backup data may be somewhat different from the data of the desired recovery time due to limited resources and the like. For example, if tapes are used to store backup data, the tapes are reused over and over again since the number of tapes is limited. At this time, if a backup data is acquired every day at 0:00 and data destruction occurs at 23:00, the latest backup data using the backup data in the tape is 23 hours before the desired recovery time. In other words, the recovery point intervals are too long.
In order to solve this problem, an art is provided in which journal data of the normal operation data are acquired to recover lost data (for example, refer to US Patent Application Publication 2005/0015416 A1, hereinafter referred to as patent document 1). According to patent document 1, journaling of write data to the storages is performed to create a volume having the data of the volume immediately before the journaling (hereinafter referred to as “base volume”), to enable recovery of data at any point in time by applying the journal of a desired point in time to the base volume.
The journal may be defined as an after journal, which is the journal of the data after the update of write data, and a before journal, which is the journal of the data before the update of write data (refer for example to patent document 1). By acquiring after journals and before journals, and by applying the after journals and before journals to a volume taken at some timing, it becomes possible to return the volume data to an older data or to update the volume data to a newer data. Thus, data of any return timing can be acquired and recovery data can be created using the journal data with flexibility.
The conventional method described above has the following drawbacks. Since data is added to the journal every time a write data is acquired, if the amount of write data becomes excessive, the usage of the journal data storage area may become 100%. Then, the journal data may overflow from the storage area so that the latest journal data cannot be added, or the journal data may overwrite necessary data so that old journal data is deleted.
According to patent document 1, a snapshot is created using the base volume and journal data before the data overflows, and the old journal data used for creating the snapshot can be deleted after the snapshot is created. However, according to the art of patent document 1, the processing costs related to creating snapshots are great if there are much journal data, and the number of snapshots and as a result the number of volumes become excessive, using up much management resources within the storages. Therefore, it becomes necessary to set up appropriate capacities of the areas for storing the journal data (hereinafter referred to as journal volume) and the timing for creating the snapshots, and present the result to the user.
Furthermore, the user must designate the capacity of the journal volume by considering the access status of the hosts and usable resources, and the user must designate the recovery demand by considering the timing for returning to the recovery point and the time required for recovery. Therefore, it is difficult for the user to estimate the required capacity.